A 32P-postlabelling assay was developed for the analysis of adducts arising from the reaction of 2'-deoxyguanosine-3'-monophosphate with acetaldehyde, the primary oxidative metabolite of ethanol. The 32P-postlabelling reaction was optimized by testing various parameters such as the kinetics of phosphorylation by T4 polynucleotide kinase, substrate-concentration-dependent labelling efficiency and the concentration of the various ingredients of the phosphorylation reaction. The sensitivity to 3'-monophosphate dephosphorylation activity of nuclease P1 was also studied. Three stable adducts were separated by reversed-phase HPLC. The major stable adduct was structurally characterized and identified as N2-ethyl-2'-deoxyguanosine and could be detected, after reduction with NaBH4 or a mixture of ascorbic acid and GSH, in calf thymus DNA samples that had been reacted in vitro with acetaldehyde. DNA adducts were isolated after enzymatic digestion to mononucleotides followed by nuclease P1 digestion of normal nucleotides. The average levels of acetaldehyde-DNA adducts detected in these samples were 12.1 +/- 2.3 (n = 17) and 4.9 +/- 0.9 (n = 9) adducts/10(7) nucleotides after reduction with NaBH4, or ascorbic acid and GSH respectively. The 32P-postlabelling method was further validated by the detection of acetaldehyde adducts in liver DNA from mice treated with ethanol. The average concentration of the adducts detected in these animals was 1.5 +/- 0.8 (n = 7) adducts/10(8) nucleotides, as analyzed by reversed-phase HPLC with online detection of radioactivity.